Size-Distribution Scaling in Clusters of Allelomimetic Agents

Dranreb Earl Juanico and Caesar Saloma

National Institute of Physics
University of the Philippines, Diliman
dojuanico@up.edu.ph

Extended Abstract
This article presents a simple, yet general, model that
explains the scaling properties observed in real cluster-
size distributions.
The model defines a threshold response for an agent
expressed by
( )

>
=
c
c
P
if , 0
if ,
, (1)
Equation (1) gives the probability that an agent actu-
ates depending on its stimulus level and
c
is some
threshold value. Such characteristic is akin to neuronal
stimulation. Furthermore, equation (1) states that an
agent may override the consequence of >
c
by a
complementary choice of not performing an action.
This is reflected by which may be less than 1. Equa-
tion (1) therefore incorporates the ability of an agent to
decide. Agents are distributed in a d-dimensional lat-
tice of length L. A fraction p of the agent population is
designated to be unresponsive. Unresponsive agents are
characterized by =0 such that P(,)=0 even if >
c
.
These agents may be thought of as impurities in the
lattice. The value of p is varied between 0 and 1.
The lattice is constantly bombarded by external stim-
uli (e.g., environmental conditions or public informa-
tion in the form of advertisements, and the like). The
consequence of this is that at each time step of the nu-
merical experiment, +1 for all agents (i.e., stimu-
lus levels are raised by a unit).
An agent is then randomly chosen, say an agent at
cell i,j in a d=2 lattice, to behave according to (1) with
=1. Hence, if (i,j) exceeds
c
(i,j) then this agent out-
puts a particular action A. Allow us to distinguish this
chosen agent as a harbinger. The harbinger is the ini-
tiator of an action. By the performance of A the har-
bingers stimulus level decreases: (i,j)(i,j)2d, as
though releasing tension. Meanwhile, the stimulus
level of each of the harbingers 2d nearest-neighbors is
increased by a unit due to their observance of A. For a
responsive neighbor, sufficient stimulation (>
c
) result-
ing from the observance of A makes it actuate A with a
probability . Let us assume that is of a particular
value , wherein is defined as the allelomimesis in-
dex. In contrast, any amount of stimulation brought
about by the observance of A will have no effect on an
unresponsive neighbor.
The harbingers neighbors in turn pass around the in-
formation to their corresponding neighbors by actuat-
ing A. The dynamics of action propagation can be
summarized as two fundamental processes: (i) the se-
lection of a harbinger that initiates an action, and (ii)
propagation of action through nearest-neighbor connec-
tions. Process (ii) is repeated until the action initiated
by the harbinger ceases to propagate. All agents that
actuate or have actuated A are considered to belong to
a cluster and the total number of these agents corre-
sponds to the size s of the cluster. The resulting distri-
bution f(s) is highly-dependent on and p.

Figure 1 Clusters that result for =1and p=0

Figure 1 illustrates the morphology of clusters result-
ing from the model. These clusters apparently exhibit
fractal structures that were produced by models of ur-
ban growth through diffusion-limited aggregation [1].
The model is able to fit cluster-size distributions of
actual clustering systems in nature. This is done by
simply tuning the values of and p. Our significant
finding is that have high-values for animals implying
that allelomimesis is strongly expressed whereas ~0.3
for human beings. Details of the data analysis method
utilized in relating the model with data gathered from
multifarious real clustering is discussed in Ref. 2.
References
[1] H. Makse, S. Havlin and H.E. Stanley, Modeling
urban growth patterns, Nature 377, 608, 1995.
[2] In Sec. 4.5 of D.E. J uanico, Agent-Based Modeling
of Self-Organized Clustering in Nature Based on Al-
lelomimesis, Masters Thesis, College of Science, Uni-
versity of the Philippines Diliman, 2004.
Size-Distribution Scaling in Clusters of Allelomimetic Agents
Dranreb Earl J uanico and Caesar Saloma
National Institute of Physics
University of the Philippines, Diliman
dojuanico@up.edu.ph

Abstract
The allelomimesis clustering model is based on
only two parameters: a local parameter that
represents the probability of nearest-neighbor
copying and a global parameter p that represents the
fraction of unresponsive agents. The model results
into the formation of clusters of agents, the sizes of
which obey a distribution that is determined by the
values of and p. Several experimental data are
fitted by tuning the two parameters. In particular, the
significance of the value of that corresponds to an
experimental data is discussed and is justified
according to behavioral context. Recommendations
for possible extensions of the model are also
enumerated.
1. Introduction
The penultimate hallmark of complex systems is
the principle of emergence macroscopic regularity
of the system arising from apparently irregular
(disordered) microscopic interactions between the
systems constituents. Among the most commonly
observed of these emergent properties is clustering.
Clusters in nature exist in different sizes. An
interesting and somewhat unexpected observation is
the regularity of the statistical distribution of cluster
sizes, which generally obeys what is known as a
power-law. If we denote by s the size of a cluster and
by f(s) the frequency of occurrence of s (i.e., the
number of times that a particular value of s is tallied),
then one would witness that the plot of f(s) versus s in
double logarithmic scale is a decreasing straight line.
The slope of this line corresponds to the exponent of
the power law. Let us refer to the value of this slope
as .
Power-law behavior is a well-known result in the
area of complex systems. This article presents a
simple, yet general, mechanism that leads to this
power-law behavior.
In section 2, a phenomenological model based on
allelomimetic behavior is discussed. Allelomimesis is
the tendency of individuals to imitate its neighbors;
hence, allelomimetic behavior is the best candidate as
local interaction that could lead to the formation of
clusters. The results of the model are discussed in
Section 3. A comparison of these results with several
cluster systems found in nature is also presented. A few
recommendations for further extension of this study are
pointed out in Section 4.
2. The Clustering Model
2.1 Behavior of a single agent
The probability that an agent performs a certain
action is described as follows:
( )

>
=
c
c
P
if , 0
if , 1
(1)
wherein is the total stimulus received by an agent
(both from its environment and its neighboring agents)
and
c
is an arbitrary threshold stimulus level. Equation
(1) means that an agent actuates if its stimulus level
exceeds the threshold. Such characteristic is akin to
neuronal stimulation [1].
In reality, however, an agent may override the
consequence of >
c
by a complementary choice of
not performing an action. Let us assume that the
probability the agent will not actuate if >
c
is 1.
Hence, (1) may be rewritten in the following form:
( )

>
=
c
c
P
if , 0
if ,
, (2)
Equation (2) incorporates the ability of an agent to
decide.
Generally, and
c
may vary among agents. But to
make the model as simple as possible, it is assumed
that is a mean value over a population of agents,
hence, is a constant with respect to a particular agent
population. However, is allowed to vary between
different populations. On the other hand, two cases are
considered in assigning the value of
c
it is either
fixed (
c
=4) for all agents or it varies within a range
(2
c
16) among agents.

2.2 Lattice of agents
Agents are distributed in a d-dimensional lattice of
length L consisting of L
d
discrete cells. A cell may only
accommodate a single agent. Thus, L
d
also corresponds
to the size of the agent population.
A fraction p of the agent population is designated to
be unresponsive. Unresponsive agents are
characterized by =0 such that P(,)=0 even if
>
c
. These agents may be thought of as impurities
in the lattice. The value of p is varied between 0 and
1.

2.3 Dynamics of action propagation
The lattice is constantly bombarded by external
stimuli (e.g., environmental conditions or public
information in the form of advertisements, and the
like). The consequence of this is that at each time
step of the numerical experiment, +1 for all
agents (i.e., stimulus levels are raised by a unit).
An agent is then randomly chosen, say an agent at
cell i,j in a d=2 lattice, to behave according to (1).
Hence, if (i,j) exceeds
c
(i,j) then this agent outputs
a particular action A. Otherwise, nothing happens and
another agent is randomly chosen. Allow us to
distinguish this chosen agent as a harbinger. The
harbinger is the initiator of an action. Once a
harbinger is selected, the bombardment of external
stimuli is momentarily paused to allow us to focus on
the consequence of the harbingers action to the
entire lattice. Furthermore, we assume that there can
only be one harbinger at a time but any agent is a
potential harbinger at any given time.
By the performance of A the harbingers stimulus
level decreases: (i,j)(i,j)2d, as though releasing
tension. Meanwhile, the stimulus level of each of the
harbingers 2d nearest-neighbors is increased by a
unit due to their observance of A. These neighbors
behave according to (2) to decide whether or not to
mimic the harbinger and actuate A. For a responsive
neighbor, sufficient stimulation (>
c
) resulting from
the observance of A makes it actuate A with a
probability . Let us assume that is of a particular
value . In contrast, any amount of stimulation
brought about by the observance of A will have no
effect on an unresponsive neighbor.
The harbingers neighbors in turn pass around the
information to their corresponding neighbors by
actuating A. This propagation continues up to the last
agent that performs A without influencing its
corresponding nearest neighbors.
The parameter is defined as the allelomimesis
index. Its value is tuned between 0 and 1. On one
hand, =0 is equivalent to setting p=1, i.e. all agents
in the population are unresponsive to their neighbors;
hence, non-copying or non-allelomimetic. On the
other hand, =1 implies a highly-allelomimetic
population of agents wherein imitation of neighbors
is a big factor that promotes clustering.
The dynamics of action propagation can be
summarized as two fundamental processes: (i) the
selection of a harbinger that initiates an action, and (ii)
propagation of action through nearest-neighbor
connections.

2.4 Cluster and cluster size
Process (ii) is repeated until the action initiated by the
harbinger ceases to propagate. All agents that actuate
or have actuated A are considered to belong to a cluster
and the total number of these agents corresponds to the
size of the cluster s. Subsequently, the bombardment of
external stimuli is resumed for the proceeding time
step. Process (i) results to the initiation of another
action and process (ii) propagates this action through
the lattice, hence, establishing the formation of another
cluster. By repeating processes (i) and (ii) over several
time steps, one generates different values of s. This
allows one to deduce the statistical distribution f(s).
3. Results and Discussion
3.1 Numerical simulations
Figure 1 illustrates the morphology of clusters at
different settings of the parameters and p. These
clusters apparently exhibit fractal structures resembling
those that were produced by models of urban growth
through diffusion-limited aggregation [2].
Let us first deal with the effect of varying by
setting p =0. Figure 2 plots the power-law cluster size
distribution (CSD) in double logarithmic scale for
different values of . Notice how the lines steepen with
increasing value of , indicating that the scaling
exponent is negatively correlated with the parameter
. It is expected that as 0, consistent with a
dirac-delta CSD centered at s=1 for =0 (i.e., no
clusters are formed).
To show the effect of the parameter p on the CSD,
we fix to a value of 1. Figure 3 exhibits a distortion
of the CSD at large values of s. The degree of such
distortion intensifies with increasing p.
Considering that thresholds
c
may vary among
agents, we compare the set of CSDs (with different
and p =0) for the case wherein 2
c
16 with the set
for which
c
=4. Figure 4 plots the CSDs as data points
in the former case and as broken lines in the latter case.
There is no observable difference and this implies that
the exact value of
c
of an agent does not affect the
CSD. Hence, the CSD is robust to variations of
c

within an agent population.

3.2 Comparison with data for real systems
We fit our model to different CSDs taken from
experimental observations of actual clusters. The
goodness-of-fit is measured in terms of the mean
square error (MSE) between the data and the curve
generated from the model. Figures 5 and 6 present
data on four animal systems and four distinct human
cluster systems, respectively.
Remarkably, is high for animal systems (except
for Serengeti lions) implying that allelomimesis is
strongly expressed in animals. According to Wagner
and Danchin, conspecific copying (or
allelomimesis) is a ubiquitous mechanism behind the
formation of aggregates such as leks and colonies [3].
Bonabeau and Dagorn showed that biosocial
attraction (another form of allelomimesis) promotes
schooling in fishes [4]. Parrish and Keshet further
proposed that allelomimesis is a generic mechanism
that maintains the cluster as a cohesive unit [5].
Indeed, ecological evidence for a high value of in
the animal kingdom is compelling. The seemingly
low (=0.1) for Serengeti lions is compensated by
the high-value of p. Such disparity may be explained
by the fact that the lions that were observed by
Schaller were nomadic [6]. This means that they are
likely to wander alone or in small groups, hence,
even though lions may be considered highly-social
(high p), being nomadic disrupts the information flow
between lions resulting in low . A comprehensive
discussion of the deduced values of and p is found
in Ref. 7.
The value of for human cluster systems is low as
compared to animal systems, implying that
allelomimesis is only moderately expressed in human
beings. This can be justified by considering that
humans are generally more highly cognitive than
animals, which consequently overrides their
instinctive tendency to be allelomimetic.
Furthermore, telecommunication technology (which
only humans are capable of) diminishes the
requirement of information transfer through nearest-
neighbor connections such as allelomimesis.
Interestingly, is not significantly different among
distinct human cluster systems (0.3). This result is
quite expected because even though we consider
clusters of cities, of households or of employees to be
distinct from one another, one fact remains common
between them these systems are all made up of
human beings. It would be worthwhile to investigate
the origin of such seemingly universal value of
from a psychological point of view.
4. Recommendations
The model due to its inherent simplicity has cut down
on details as much as possible so that it can be
considered generic, hence, applicable to a wide variety
of systems. Here, we suggest some minor points of
modification to allow a more realistic description.
The stimulation on an agent due to constant
bombardment of external factors may not necessarily
be equal to unity (i.e., +1). It can be expressed as
+, where represents a positive Poisson number
that appropriately describes the time fluctuation of the
amount of external stimuli. Furthermore, subsequent
stimulation of neighboring agents may not necessarily
decrease the stimulus level of the harbinger by an
amount that is equal to the number of its nearest
neighbors. That is, one can write , where is a
positive number derived from a Gaussian or a Binomial
probability distribution. It follows that the ensuing
stimulation of responsive neighbors that observe the
actuation of the harbinger can be expressed as

n
+
n
where the subscript n represents the nearest-
neighbor and
n

n
=. Note here that
n
could either be
positive or negative, implying that the stimulation is
excitatory or inhibitory, respectively [1].
5. Conclusion
A simple model of cluster formation is proposed to
explain the cluster size distribution observed for
various cluster systems in nature. The model consists
of two mutually independent parameters, namely and
p. The value of represents the probability that an
agent mimics the action of its nearest-neighbors
whereas p is the fraction of unresponsive agents that
characterizes the particular agent population. Resulting
CSDs are highly-dependent on and p.
The model fits into experimental data corresponding
to various cluster systems in nature. High value of
generally characterizes animal systems whereas ~0.3
distinguishes human cluster systems.

Figure 1 Clusters that result for =1and p=0.

Figure 2 CSD for different values of at p=0.

Figure 3 CSD at =0 and different values of p.

Figure 4 Comparison between imposing uniform
threshold (
c
=4) and random threshold (2
c
16)
for different values of at p=0.

Figure 5 Fitting the model to four different
animal systems. A Spotted dolphin, Stenella
attenuata (=0.75, p=0.3, MSE=0.00381); B West
Indian manatee, Trichecus manatus (=1, p=0.45,
MSE=0.00034); C Wasp, Ropalidia fasciata
(=0.75, p=0.35, MSE=0.00059); D Serengeti lion
Panthera leo (=0.1, p=0, MSE=0.08661).
A
B
C
D

Figure 6 Fitting the model to four distinct
human cluster systems. A Urban agglomerations
of India, 1991 (=0.32, p=0, MSE=0.00394); B
Major cities of Japan, 1994 (=28, p=0,
MSE=0.00021); C Households/barrios of Metro
Cebu, 2001 (=0.31, p=0.25, MSE=0.00179); D
Firms/clusters of employees of U.S., 1997 (=0.32,
p=0, MSE=6.16504).

A
References
[1] P.S. Churchland and T. Sejnowski, The
Computational Brain, MIT Press, Cambridge, 1992.
[2] H. Makse, S. Havlin and H.E. Stanley,
Modeling urban growth patterns, Nature 377, 608,
1995.
[3] R.H. Wagner, E. Danchin, Conspecific copying:
a general mechanism of social aggregation, Anim.
Behav. 65, 405, 2003.
[4] E. Bonabeau and L. Dagorn, Possible
universality in the size distribution of fish schools,
Phys. Rev. E 51, 5220, 1995.
B
[5] J .K. Parrish and L. Edelstein-Keshet, From
individuals to aggregations: Complexity,
epiphenomena, and evolutionary trade-offs of animal
aggregation, Science 284, 99, 1999.
[6] G.B. Schaller, The Serengeti Lion: A Study of
Predator-Prey Relations, Univ. of Chicago Press,
1972.
[7] In Section 4.5 of D.E. Juanico, Agent-Based
Modeling of Self-Organized Clustering in Nature
Based on Allelomimesis, Masters Thesis, College of
Science, University of the Philippines Diliman,
2004.
C
D